EP1599833A1 - Image segmentation by assigning classes to adaptive mesh primitives - Google Patents

Image segmentation by assigning classes to adaptive mesh primitives

Info

Publication number
EP1599833A1
EP1599833A1 EP04709288A EP04709288A EP1599833A1 EP 1599833 A1 EP1599833 A1 EP 1599833A1 EP 04709288 A EP04709288 A EP 04709288A EP 04709288 A EP04709288 A EP 04709288A EP 1599833 A1 EP1599833 A1 EP 1599833A1
Authority
EP
European Patent Office
Prior art keywords
mesh
classes
class
primitives
probability
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
EP04709288A
Other languages
German (de)
French (fr)
Other versions
EP1599833B1 (en
Inventor
Jens Philips I.. P. & Standards. GmbH VON BERG
Michael Philips I. P. & Standards GmbH KAUS
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Philips Intellectual Property and Standards GmbH
Koninklijke Philips NV
Original Assignee
Philips Intellectual Property and Standards GmbH
Koninklijke Philips Electronics NV
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Philips Intellectual Property and Standards GmbH, Koninklijke Philips Electronics NV filed Critical Philips Intellectual Property and Standards GmbH
Priority to EP04709288A priority Critical patent/EP1599833B1/en
Publication of EP1599833A1 publication Critical patent/EP1599833A1/en
Application granted granted Critical
Publication of EP1599833B1 publication Critical patent/EP1599833B1/en
Anticipated expiration legal-status Critical
Expired - Lifetime legal-status Critical Current

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/0002—Inspection of images, e.g. flaw detection
    • G06T7/0012—Biomedical image inspection
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/12—Edge-based segmentation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/143—Segmentation; Edge detection involving probabilistic approaches, e.g. Markov random field [MRF] modelling
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/149—Segmentation; Edge detection involving deformable models, e.g. active contour models
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/10—Image acquisition modality
    • G06T2207/10072—Tomographic images
    • G06T2207/10081—Computed x-ray tomography [CT]
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/30—Subject of image; Context of image processing
    • G06T2207/30004—Biomedical image processing

Definitions

  • the present invention generally relates to image segmentation.
  • the present invention relates to the quantification of geometric functional parameters of a selected region in a multi-dimensional dataset.
  • the present invention relates to a method of segmenting a selected region from a multi-dimensional dataset, to a method of assigning classes to mesh primitives of an adapted mesh for segmenting an object from an image, to a method of generating a shape model for segmentation of a object from a multi-dimensional dataset, to an image processing device and to a computer program for an image processing device for assigning classes to mesh primitives of an adaptive mesh for segmenting an object from the image.
  • Image segmentation generally concerns selection and/or segmentation of a selected part of a dataset.
  • Such a dataset notably represents image information of an imaged object and the selected part relates to a specific part or object of the image.
  • the dataset is in general a multi-dimensional dataset that assigns data values to positions in a multi-dimensional geometrical space.
  • such datasets can be two- dimensional or tliree-dimensional images, where the data values are pixel values, such as brightness values, gray values or color values, assigned to positions in a two- dimensional plane or a three-dimensional volume.
  • a selection of the selected region may be affected by an initialization of the segmentation.
  • the initialization may be done on the basis of the shape model.
  • a shape model represents the selected region in terms of a parameterization of a priori shape information. This a priori shape information and the positioning of the shape model in the multi-dimensional dataset determine the region to be selected.
  • the shape model is an active shape model that allows for adaptation of its translation and orientation and pose, but also of shape parameters representing smaller details of the shape of the region represented by the active shape model.
  • the adaptive mesh represents an approximate contour.
  • Such an approximate contour represents an estimate of a boundary of the selected region with respect to the remainder of the multidimensional dataset.
  • Such a contour is usually a hyper-surface in the multi-dimensional dataset, such as a surface or a curve.
  • the contour has a dimension which is preferably smaller than the dimensionality of the multi-dimensional dataset.
  • the approximate contour approaches the boundary of the selected region.
  • US 2002/0184470 Al discloses such a method, where a shape model is set up representing the general outline of the selected region, an adaptive mesh is set up representing an approximate contour of the selected region, and wherein the adaptive mesh is initialized on the basis of the shape model and deformed in dependence on the shape model and on feature information of the selected region.
  • the mesh adaptation is performed on the basis of optimizing a value of an energy function, wherein the energy function has an internal energy contribution, which depends on the shape model and an external energy contribution which depends on feature information of the selected region and the actual configuration of the adaptive mesh.
  • the term "energy” is not related to physical energy, but is used because the concept of a deformable adaptive mesh shows strong resemblance to a string of masses (vertices) interconnected by springs (i.e. bonds), which move under the influence of a force field and strive towards a stable situation of minimum energy.
  • the external energy pertains to the external force field applied to the vertices and the internal energy pertains to the interaction between the vertices.
  • the internal energy relates to a deviation of a primitive of the adaptive mesh from its original position in the shape model during the adaptation of the mesh to the selected region.
  • the external energy relates to the deviation of this surface primitive of the adaptive mesh to the boundary of the selected region.
  • Such segmentation methods are used to derive, for example, models of organs or bones from the volumetric image data such as, for example, CT, MR, or US images.
  • the geometric models are required for a variety of medical applications, or generally in the field of pattern recognition.
  • cardiac diagnosis where geometric models of the ventricles and the myocard of the heart are required for perfusion analysis, or motion analysis and computation of the ejection fraction.
  • radiotherapy planning where the segmentation of multiple organs and bones in, for example, the prostate region (femur heads, rectum, prostate, bladder) is necessary for the diagnosis and/or for the determination of treatment parameters. Due to the size and complexity of such large datasets, automatic tools for quantitative measurements are important for fast, accurate and reproducible analysis of such datasets in daily routine. However, the known methods are often confound by spatially varying features around the selected region.
  • this object may be solved with a method of segmenting a selected region from a multidimensional dataset, wherein an adaptive mesh comprising a plurality of mesh primitives is to be adapted to the selected region.
  • data is sampled from the multi-dimensional dataset.
  • a probability is estimated that a mesh primitive of the plurality of mesh primitives belongs to a class of a plurality of classes by using the sample data from the multi-dimensional dataset by means of an estimator.
  • an energy function corresponding to the specific class of the plurality of classes is applied.
  • this exemplary embodiment of the present invention a method is provided coping with the case where there is no strong agreement in the spatial extent of tissue interfaces around the organs.
  • the method according to this exemplary embodiment copes with spatially varying features around the selected region.
  • this exemplary embodiment of the present invention provides for a rapid, accurate and robust method enabling the use thereof in a medical environment in daily routine.
  • the estimator is a Bayesian conditional estimator. This allows for a very rapid execution of the method according to the present invention with a high accuracy.
  • the plurality of classes relate to different features of the selected region, allowing for an accurate segmentation of particular features of the selected region.
  • Another exemplary embodiment of the present invention as set forth in claim 4 relates to a method of assigning classes to mesh primitives of an adaptive mesh for segmenting an object from an image, where a feature probe is used to assign a class to a mesh primitive.
  • an accurate and dependable classification with respect to features of the selected region can be made, even if there are spatially varying features around the object.
  • an estimate is used for the classification which is a Bayesian conditional estimator.
  • Another exemplary embodiment of the present invention as set forth in claim 6 relates to a method of generating a shape model for segmentation of an object from a multi-dimensional dataset, which allows for an adaptive shape model.
  • primitives of the adaptive mesh of the shape model are not fixedly assigned to classes related to features of the object, but the class membership of a mesh primitive may change during the adaptation, which allows for a very accurate segmentation when the shape model is used for adaptation.
  • an a posteriori probability that a feature vector belongs to a class is determined on the basis of Bayes rules. This provides for a very secure and accurate while flexible segmentation.
  • Claims 8 and 9 relate to image processing devices according to exemplary embodiments of the present invention, which allow for an improved segmentation.
  • the present invention relates also to a computer program for an image processing device for assigning classes to mesh primitives of an adaptive mesh for segmenting an object from the image.
  • the computer program according to the present invention is defined in claim 10.
  • the computer program according to the present invention is preferably loaded into a working memory of a data processor.
  • the data processor is thus equipped to carry out the method of the invention.
  • the computer program may be stored on a computer readable medium, such as a CD-Rom.
  • the computer program may also be presented over a network such as the worldwide web and can be downloaded into the working memory of a data processor from such a network.
  • an estimator is applied to assign feature classes to mesh primitives that takes the actual feature probe from the unseen image into account, calculates an a posteriori probability that the mesh primitive belongs to a given class and uses the according external energy of that class.
  • Fig. 1 shows a schematic representation of an image processing device according to an exemplary embodiment of the present invention adapted to execute a method according to an exemplary embodiment of the present invention.
  • Fig. 2 shows a flow-chart of an exemplary embodiment of a method according to the present invention.
  • Fig. 3 shows a flow-chart of another exemplary embodiment of a method according to the present invention.
  • Fig. 4 shows segmentation results of a ventricle segmented according to the present invention.
  • Fig. 5 shows further segmentation results of a ventricle segmented according to the present invention.
  • Fig. 6 shows surface measures of two individual left ventricles.
  • Fig. 1 shows a simplified schematic representation of an embodiment of an image processing devices in accordance with the present invention.
  • a central processing unit (CPU) or image processor 1 for adapting a deformable model surface, i.e. a surface mesh of a shape model to a selected region by mesh adaptation.
  • the image processor 1 is connected to a memory 2 for storing multidimensional datasets such as, for example, images.
  • the image processor 1 may be connected via a bus system 3 to a plurality of periphery devices or input/output devices, which are not depicted in Fig. 1.
  • the image processor 1 may be connected to a MR device, a CT device, to a plotter or a printer or the like via bus system 3.
  • the image processor 1 is connected to a display such as a computer screen 4 for outputting segmentation results or information.
  • a keyboard 5 is provided connected to the image processor 1 , by which a user or operator may interact with the image processor 1 or may input data necessary or desired for segmentation.
  • Figs. 2 and 3 show flow-charts of exemplary embodiments of methods for operating the image processing device according to the present invention as shown in Fig. 1.
  • Fig. 2 relates to a method for generating a shape model for segmenting an object from a multi-dimensional dataset
  • the method depicted in Fig. 3 relates to the adaptation of the shape model for segmenting an object from an unseen image.
  • step SI the method continues to step S2 where an a priori probability that a mesh primitive P ! of the plurality of mesh primitives of the mesh of the shape model belongs to a class C, of a plurality of classes relating to features of the object.
  • the a priori probability can be written as P, (C,).
  • P, (C,) may, according to an exemplary embodiment of the present invention, be directly derived from a frequency of occurrence in the training set during the training, i.e. the formation of the shape model, by applying the shape model to training sets. Then, the method continues to step S3, where a class conditional probability density function is estimated for a feature vector v at a position in the multi-dimensional dataset for the plurality of classes by using data from a training model.
  • the class conditional probability density function for a given feature vector v is estimated for each C j as p(v
  • a parametric density function such as for example a multidimensional Gaussian function is used to model this data from the training set. Now, training is completed and the model is built.
  • step S4 an a posteriori probability that the measured feature factor v at the position i is of class C j is determined.
  • v) can be determined as follows by applying the Bayes' rules as for example described in R.O. Duda et al, "Patent classification and scene analysis” John Wiley&Sons, 1973, which is hereby incorporated by reference in its entirety:
  • Steps S2 to S4 may be repeated for the feature vector v at each position i and a decision may be made with respect to a best fitting class of the plurality of classes for the feature vector v at the position i on the basis of the highest a posteriori probability.
  • step S5 After the determination of the best fitting class for feature vector v at the position i, the method continues to step S5, where it ends.
  • step SI 5 the method continues to step SI 5.
  • a conditional decision process is applied, where the measured feature probe from the unseen image is incorporated into the decision process.
  • a class membership of a mesh primitive Pj may change during the adaptation and thus the local optimization goal.
  • FIG. 4 shows slice images of the ventricles
  • Fig. 5 shows three-dimensional representations of the ventricles.
  • the borders of tissue interface types of the left heart ventricle may vary between subjects and thus a position in the surface model once belongs to the interface of the lung, in Fig. 4 the upper ventricle and in Fig. 5 the left ventricle and once to the right ventricle as shown in the lower portion of Fig. 4 and on the right portion of Fig. 5.
  • a statistical model that only represents a priori probabilities of tissue interface will always decide for the same tissue interface (namely the most likely one) at a given position.
  • a feature probe from the unseen image is incorporated in the decision process and decides which tissue interface it belongs to.
  • tissue types of minor a priori probability can be identified if the evidence from the probe is sufficiently high.
  • Fig. 6 shows a surface mesh of two individual ventricles viewed from the apex and flattened.
  • the surface is separated into tissue interface classes I-r, 1-1 and I-f, as indicated by the bold lines in Fig. 6, indicating boundaries between the classes.
  • tissue interface classes I-r, 1-1 and I-f there are three relevant tissue interface classes, namely, the interface of the right ventricle I-r, the interface to the lung 1-1 and the interface to epicardial fat I-f.
  • a profile of the interface I-f may look like the I- r interface, also can be found in a position where there is usually I-i to be found and hardly any I-r.
  • a statistical model for a given 1-1 dominated position on the surface that only takes mean values of the uni -modal distribution for all subjects into account, will only be able to decide on one choice and thus may cause problems to represent that above mentioned variability in the learning set.
  • the occurrence of epicardial fat in an unseen image at that position will therefore confound the method, although it is to be expected from the learning set to a certain a priori probability.
  • the adaptive conditional estimator is used, allowing that the actual measured feature probe can be classified by a Bayes decision theory that gives the hint at what interface class occurs in the unseen image.
  • a class of tissue interfaces may be varyingly observed in the training set at a given mesh position, advantageously, its occurrence in an unseen image can be detected, since the measured feature probe is taken into account.
  • deformable models according to the present invention using this kind of estimator according to the present invention are more flexible, robust and accurate.
  • the method according to the present invention of adaptive conditional varying surface features still finds evidence for a given class where known methods failed.
  • This a posteriori probability is the probability that a feature vector v measured at the position i belongs to the best fitting class C b , v) . At low confidence values no external energy may be used for the position i, or the external energy contribution may be at least reduced there.
  • a risk function as described in R.O. Duda et al, "Patent classification and scene analysis” John Wiley&Sons, 1973), which is hereby incorporated by reference in its entirety, in order to model the impact of a false decision. It may depend on the kind of application which risk function should be used. For example, in radiation therapy planning, a mis-classification leading to exposure of sensitive tissue needs a higher risk factor than other tissue classes.
  • the present invention is applicable to any multi-dimensional datasets, where selected regions are to be segmented. Prominent applications may also include functional diagnostics of soft tissue organs, like the heart, brain or liver from MR and CT and radiotherapy planning (bones and soft tissue from CT).

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Software Systems (AREA)
  • Health & Medical Sciences (AREA)
  • General Health & Medical Sciences (AREA)
  • Medical Informatics (AREA)
  • Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
  • Radiology & Medical Imaging (AREA)
  • Quality & Reliability (AREA)
  • Probability & Statistics with Applications (AREA)
  • Image Analysis (AREA)
  • Image Generation (AREA)

Abstract

Quantification of metric or functional parameters often requires image segmentation. A crucial part of such method is the model of the surface characteristics of the object of interest (features), which drives the deformable surface towards the object boundary in the image. According to the present invention, sections of the mesh are assigned to different classes for different features. According to the present invention, the assignment of mesh sections to the classes is adapted by using actual feature information from the unseen image. Advantageously, this allows for an adaptation of the feature category to which the mesh section is assigned and thereby allows an improved segmentation of the object.

Description

IMAGE SEGMENTATION BY ASSIGNING CLASSES TO ADAPTIVE MESH PRIMITIVES
The present invention generally relates to image segmentation. In more detail, the present invention relates to the quantification of geometric functional parameters of a selected region in a multi-dimensional dataset. In particular, the present invention relates to a method of segmenting a selected region from a multi-dimensional dataset, to a method of assigning classes to mesh primitives of an adapted mesh for segmenting an object from an image, to a method of generating a shape model for segmentation of a object from a multi-dimensional dataset, to an image processing device and to a computer program for an image processing device for assigning classes to mesh primitives of an adaptive mesh for segmenting an object from the image. Image segmentation generally concerns selection and/or segmentation of a selected part of a dataset. Such a dataset notably represents image information of an imaged object and the selected part relates to a specific part or object of the image. The dataset is in general a multi-dimensional dataset that assigns data values to positions in a multi-dimensional geometrical space. In particular, such datasets can be two- dimensional or tliree-dimensional images, where the data values are pixel values, such as brightness values, gray values or color values, assigned to positions in a two- dimensional plane or a three-dimensional volume.
A selection of the selected region may be affected by an initialization of the segmentation. In particular, the initialization may be done on the basis of the shape model. Such a shape model represents the selected region in terms of a parameterization of a priori shape information. This a priori shape information and the positioning of the shape model in the multi-dimensional dataset determine the region to be selected. In particular, the shape model is an active shape model that allows for adaptation of its translation and orientation and pose, but also of shape parameters representing smaller details of the shape of the region represented by the active shape model. The adaptive mesh represents an approximate contour. Such an approximate contour represents an estimate of a boundary of the selected region with respect to the remainder of the multidimensional dataset. Such a contour is usually a hyper-surface in the multi-dimensional dataset, such as a surface or a curve.
The contour has a dimension which is preferably smaller than the dimensionality of the multi-dimensional dataset. During the deformation of the adaptive mesh, the approximate contour approaches the boundary of the selected region.
US 2002/0184470 Al discloses such a method, where a shape model is set up representing the general outline of the selected region, an adaptive mesh is set up representing an approximate contour of the selected region, and wherein the adaptive mesh is initialized on the basis of the shape model and deformed in dependence on the shape model and on feature information of the selected region. In particular, the mesh adaptation is performed on the basis of optimizing a value of an energy function, wherein the energy function has an internal energy contribution, which depends on the shape model and an external energy contribution which depends on feature information of the selected region and the actual configuration of the adaptive mesh.
In this respect it has to be noted that the term "energy" is not related to physical energy, but is used because the concept of a deformable adaptive mesh shows strong resemblance to a string of masses (vertices) interconnected by springs (i.e. bonds), which move under the influence of a force field and strive towards a stable situation of minimum energy. In this resemblance, the external energy pertains to the external force field applied to the vertices and the internal energy pertains to the interaction between the vertices. In other words, the internal energy relates to a deviation of a primitive of the adaptive mesh from its original position in the shape model during the adaptation of the mesh to the selected region. The external energy relates to the deviation of this surface primitive of the adaptive mesh to the boundary of the selected region.
Such segmentation methods are used to derive, for example, models of organs or bones from the volumetric image data such as, for example, CT, MR, or US images. The geometric models are required for a variety of medical applications, or generally in the field of pattern recognition. For medical or clinical applications, an important example is cardiac diagnosis, where geometric models of the ventricles and the myocard of the heart are required for perfusion analysis, or motion analysis and computation of the ejection fraction. Another important clinical application is radiotherapy planning, where the segmentation of multiple organs and bones in, for example, the prostate region (femur heads, rectum, prostate, bladder) is necessary for the diagnosis and/or for the determination of treatment parameters. Due to the size and complexity of such large datasets, automatic tools for quantitative measurements are important for fast, accurate and reproducible analysis of such datasets in daily routine. However, the known methods are often confound by spatially varying features around the selected region.
It is an object of the present invention to provide for an improved image segmentation. According to an exemplary embodiment of the present invention this object may be solved with a method of segmenting a selected region from a multidimensional dataset, wherein an adaptive mesh comprising a plurality of mesh primitives is to be adapted to the selected region. For this, data is sampled from the multi-dimensional dataset. A probability is estimated that a mesh primitive of the plurality of mesh primitives belongs to a class of a plurality of classes by using the sample data from the multi-dimensional dataset by means of an estimator. To adapt the mesh primitive of the plurality of mesh primitives to the selected region, an energy function corresponding to the specific class of the plurality of classes is applied.
Advantageously, according to this exemplary embodiment of the present invention, a method is provided coping with the case where there is no strong agreement in the spatial extent of tissue interfaces around the organs. In other words, the method according to this exemplary embodiment copes with spatially varying features around the selected region. In addition to that, advantageously, this exemplary embodiment of the present invention provides for a rapid, accurate and robust method enabling the use thereof in a medical environment in daily routine.
According to another exemplary embodiment of the present invention as set forth in claim 2, the estimator is a Bayesian conditional estimator. This allows for a very rapid execution of the method according to the present invention with a high accuracy.
According to another exemplary embodiment of the present invention as set forth in claim 3, the plurality of classes relate to different features of the selected region, allowing for an accurate segmentation of particular features of the selected region.
Another exemplary embodiment of the present invention as set forth in claim 4 relates to a method of assigning classes to mesh primitives of an adaptive mesh for segmenting an object from an image, where a feature probe is used to assign a class to a mesh primitive. Advantageously according to this exemplary embodiment of the present invention an accurate and dependable classification with respect to features of the selected region can be made, even if there are spatially varying features around the object. According to an exemplary embodiment of this method as set forth in claim 5, an estimate is used for the classification which is a Bayesian conditional estimator. Another exemplary embodiment of the present invention as set forth in claim 6 relates to a method of generating a shape model for segmentation of an object from a multi-dimensional dataset, which allows for an adaptive shape model. Advantageously, primitives of the adaptive mesh of the shape model are not fixedly assigned to classes related to features of the object, but the class membership of a mesh primitive may change during the adaptation, which allows for a very accurate segmentation when the shape model is used for adaptation.
According to another exemplary embodiment of the present invention as set forth in claim 7, an a posteriori probability that a feature vector belongs to a class is determined on the basis of Bayes rules. This provides for a very secure and accurate while flexible segmentation.
Claims 8 and 9 relate to image processing devices according to exemplary embodiments of the present invention, which allow for an improved segmentation.
The present invention relates also to a computer program for an image processing device for assigning classes to mesh primitives of an adaptive mesh for segmenting an object from the image. The computer program according to the present invention is defined in claim 10. The computer program according to the present invention is preferably loaded into a working memory of a data processor. The data processor is thus equipped to carry out the method of the invention. The computer program may be stored on a computer readable medium, such as a CD-Rom. The computer program may also be presented over a network such as the worldwide web and can be downloaded into the working memory of a data processor from such a network.
It may be seen as the gist of an exemplary embodiment of the present invention that an estimator is applied to assign feature classes to mesh primitives that takes the actual feature probe from the unseen image into account, calculates an a posteriori probability that the mesh primitive belongs to a given class and uses the according external energy of that class.
These and other aspects of the present invention are apparent and will be elucidated with reference to the embodiment described hereinafter and with reference to the following drawings:
Fig. 1 shows a schematic representation of an image processing device according to an exemplary embodiment of the present invention adapted to execute a method according to an exemplary embodiment of the present invention. Fig. 2 shows a flow-chart of an exemplary embodiment of a method according to the present invention.
Fig. 3 shows a flow-chart of another exemplary embodiment of a method according to the present invention.
Fig. 4 shows segmentation results of a ventricle segmented according to the present invention.
Fig. 5 shows further segmentation results of a ventricle segmented according to the present invention.
Fig. 6 shows surface measures of two individual left ventricles.
Fig. 1 shows a simplified schematic representation of an embodiment of an image processing devices in accordance with the present invention. In Fig. 1 there is shown a central processing unit (CPU) or image processor 1 for adapting a deformable model surface, i.e. a surface mesh of a shape model to a selected region by mesh adaptation. The image processor 1 is connected to a memory 2 for storing multidimensional datasets such as, for example, images. The image processor 1 may be connected via a bus system 3 to a plurality of periphery devices or input/output devices, which are not depicted in Fig. 1. For example, the image processor 1 may be connected to a MR device, a CT device, to a plotter or a printer or the like via bus system 3. Furthermore, the image processor 1 is connected to a display such as a computer screen 4 for outputting segmentation results or information. A keyboard 5 is provided connected to the image processor 1 , by which a user or operator may interact with the image processor 1 or may input data necessary or desired for segmentation.
Figs. 2 and 3 show flow-charts of exemplary embodiments of methods for operating the image processing device according to the present invention as shown in Fig. 1. Fig. 2 relates to a method for generating a shape model for segmenting an object from a multi-dimensional dataset, whereas the method depicted in Fig. 3 relates to the adaptation of the shape model for segmenting an object from an unseen image. After the start in step SI, the method continues to step S2 where an a priori probability that a mesh primitive P! of the plurality of mesh primitives of the mesh of the shape model belongs to a class C, of a plurality of classes relating to features of the object. The a priori probability can be written as P, (C,). P, (C,) may, according to an exemplary embodiment of the present invention, be directly derived from a frequency of occurrence in the training set during the training, i.e. the formation of the shape model, by applying the shape model to training sets. Then, the method continues to step S3, where a class conditional probability density function is estimated for a feature vector v at a position in the multi-dimensional dataset for the plurality of classes by using data from a training model. The class conditional probability density function for a given feature vector v is estimated for each Cj as p(v | Cj). According to the present invention, a parametric density function such as for example a multidimensional Gaussian function is used to model this data from the training set. Now, training is completed and the model is built. Then, the method continues to step S4, where an a posteriori probability that the measured feature factor v at the position i is of class Cj is determined. The a posteriori probability P,(Cj | v) can be determined as follows by applying the Bayes' rules as for example described in R.O. Duda et al, "Patent classification and scene analysis" John Wiley&Sons, 1973, which is hereby incorporated by reference in its entirety:
where p.(v) = ∑Jp(v|c,)i»(ς). (2)
J It has to be noted that the a priori probability Pι(C,) may be expressed for the position i, while, for stating the class conditional probability density function p(v I Cj) in the model, all positions are pooled. Then, an index b of the best fitting class for the feature vector v at the position i may be calculated as follows: b (v) = arg max P (C |v) . (3 ) j
Steps S2 to S4 may be repeated for the feature vector v at each position i and a decision may be made with respect to a best fitting class of the plurality of classes for the feature vector v at the position i on the basis of the highest a posteriori probability.
After the determination of the best fitting class for feature vector v at the position i, the method continues to step S5, where it ends.
Alternatively, discriminant functions may be used for determining the index b of the best fitting class for a given feature vector v, which separate the feature space for a given model and mesh position and allow for a quick decision on b,(v). Fig. 3 shows a method for the adaptation of the shape model for segmenting an object from an unseen image. After the start at step S10, the method continues to step S 11 where a feature probe is sampled from the unseen image. Then, the method continues to step S12 where the class index k is acquired by k = b,(v). Then, the method continues to step S13, where the external energy share Ek(v) is identified, which is linked to the class identified in step S12. Then, the method continues to step SI 4, where the energy function E = Eeχt+ E,nt is optimized such that the primitives of the mesh of the shape model are adapted to the surface of the object. Then, the method continues to step SI 5. As can be seen from the methods depicted in Figs. 2 and 3, according to the present invention, instead of only calculating the external energy share Ej(v) of the mesh primitive Pj from a feature vector at the mesh position i, that was once assigned in the shape model to class j, a conditional decision process is applied, where the measured feature probe from the unseen image is incorporated into the decision process. Thus, a class membership of a mesh primitive Pj may change during the adaptation and thus the local optimization goal.
This will be further explained with reference to Figs. 4 and 5 showing images of ventricles. The arrows in Figs. 4 and 5 point at the same primitive of the same shape model after adaptation to different ventricles. The Fig. 4 shows slice images of the ventricles, whereas Fig. 5 shows three-dimensional representations of the ventricles. As can be taken from Fig. 4, the borders of tissue interface types of the left heart ventricle may vary between subjects and thus a position in the surface model once belongs to the interface of the lung, in Fig. 4 the upper ventricle and in Fig. 5 the left ventricle and once to the right ventricle as shown in the lower portion of Fig. 4 and on the right portion of Fig. 5. A statistical model that only represents a priori probabilities of tissue interface will always decide for the same tissue interface (namely the most likely one) at a given position. According to the adaptive conditional method described above, a feature probe from the unseen image is incorporated in the decision process and decides which tissue interface it belongs to. Thus, also tissue types of minor a priori probability can be identified if the evidence from the probe is sufficiently high.
Fig. 6 shows a surface mesh of two individual ventricles viewed from the apex and flattened. The surface is separated into tissue interface classes I-r, 1-1 and I-f, as indicated by the bold lines in Fig. 6, indicating boundaries between the classes. In a training set of relevant heart ventricles, as shown in Fig. 6, there are three relevant tissue interface classes, namely, the interface of the right ventricle I-r, the interface to the lung 1-1 and the interface to epicardial fat I-f. A profile of the interface I-f may look like the I- r interface, also can be found in a position where there is usually I-i to be found and hardly any I-r. A statistical model for a given 1-1 dominated position on the surface, that only takes mean values of the uni -modal distribution for all subjects into account, will only be able to decide on one choice and thus may cause problems to represent that above mentioned variability in the learning set. The occurrence of epicardial fat in an unseen image at that position will therefore confound the method, although it is to be expected from the learning set to a certain a priori probability. According to the present invention, the adaptive conditional estimator is used, allowing that the actual measured feature probe can be classified by a Bayes decision theory that gives the hint at what interface class occurs in the unseen image. Therefore, according to the present invention also a class of tissue interfaces may be varyingly observed in the training set at a given mesh position, advantageously, its occurrence in an unseen image can be detected, since the measured feature probe is taken into account. Advantageously, deformable models according to the present invention using this kind of estimator according to the present invention are more flexible, robust and accurate. Advantageously, the method according to the present invention of adaptive conditional varying surface features still finds evidence for a given class where known methods failed. It also provides a confidence measure of its decision by the a posteriori probability, which can be described with the following equation: r,W = ^( f(v)|v), (4) This a posteriori probability is the probability that a feature vector v measured at the position i belongs to the best fitting class Cb ,v) . At low confidence values no external energy may be used for the position i, or the external energy contribution may be at least reduced there.
Furthermore, it is possible to introduce a risk function as described in R.O. Duda et al, "Patent classification and scene analysis" John Wiley&Sons, 1973), which is hereby incorporated by reference in its entirety, in order to model the impact of a false decision. It may depend on the kind of application which risk function should be used. For example, in radiation therapy planning, a mis-classification leading to exposure of sensitive tissue needs a higher risk factor than other tissue classes. The present invention is applicable to any multi-dimensional datasets, where selected regions are to be segmented. Prominent applications may also include functional diagnostics of soft tissue organs, like the heart, brain or liver from MR and CT and radiotherapy planning (bones and soft tissue from CT).

Claims

CLAIMS:
1. Method of segmenting a selected region from a multi-dimensional dataset, wherein an adaptive mesh comprising a plurality of mesh primitives is to be adapted to the selected region, the method comprising the steps of: sampling data from the multi-dimensional dataset; estimating a probability that a mesh primitive of the plurality of mesh primitives belongs to a class of a plurality of classes by using the sampled data from the multi-dimensional dataset by means of an estimator; and applying an energy function corresponding to the class of the plurality of classes to adapt the mesh primitive of the plurality of mesh primitives to the selected region.
2. The method of claim 1 , wherein an adaptation of the adaptive mesh to the selected region is performed on the basis of optimising a value of an energy function; wherein the energy function has an internal energy contribution that depends on a shape model representing a general outline of the selected region; wherein the energy function has an external contribution that depends on the sampled data from the multi-dimensional dataset of the selected region and an actual configuration of the adaptive mesh; wherein the shape model is updated upon deformation of the adaptive mesh; and wherein the estimator is a Bayesian conditional estimator.
3. The method of claim 1, wherein the plurality of classes relate to different features of the selected region.
4. Method of assigning classes to mesh primitives of an adaptive mesh for segmenting an object from an image, the method comprising the steps of: estimating a probability that a mesh primitive of the mesh primitives belongs to a class of the classes by using a feature probe from the image by means of an estimator; and assigning a class to the mesh primitive on the basis of the estimated probability.
5. The method of claim 4, wherein the estimator is a Bayesian conditional estimator.
6. Method of generating a shape model for segmentation of an object from a multi-dimensional dataset, wherein the shape model includes an adaptive mesh with a plurality of mesh primitives to be adapted to the object, the method comprising the steps of
(a) determining an a priori probability that a mesh primitive of the plurality of mesh primitives belongs to a class of a plurality of classes relating to features of the object;
(b) estimating a class conditional probability density function for a feature vector at a position in the multi-dimensional dataset for the plurality of classes by using data from a training model; (c) determining an a posteriori probability that the feature vector belongs to the class of the plurality of classes.
7. The method of claim 6, wherein steps (a), (b) and (c) are repeated for each class of the plurality of classes and a decision is made with respect to a best fitting class of the plurality of classes for the feature vector at the position on the basis of the highest a posteriori probability, and wherein the a posteriori probability is determined on the basis of Bayes' rules.
8. Image processing device comprising: a memory for storing a multi- dimensional dataset; a calculation unit for segmenting a selected region from the multidimensional dataset, wherein an adaptive mesh comprising a plurality of mesh primitives is to be adapted to the selected region, the calculation unit being constructed to sample data from the multi-dimensional dataset; estimate a probability that a mesh primitive of the plurality of mesh primitives belongs to a class of a plurality of classes by using the sampled data from the multi-dimensional dataset by means of an estimator; and apply an energy function corresponding to the class of the plurality of classes to adapt the mesh primitive of the plurality of mesh primitives to the selected region.
9. Image processing device comprising: a memory for storing an image; a calculation unit for assigning classes to mesh primitives of an adaptive mesh for segmenting an object from the image, the method comprising the steps of: estimating a probability that a mesh primitive of the mesh primitives belongs to a class of the classes by using a feature probe from the image by means of an estimator; and assigning a class to the mesh primitive on the basis of the estimated probability.
10. Computer program for an image processing device for assigning classes to mesh primitives of an adaptive mesh for segmenting an object from the image, the computer program comprising computer code means for performing the following steps when the computer program is executed on the image processing device: estimating a probability that a mesh primitive of the mesh primitives belongs to a class of the classes by using a feature probe from the image by means of an estimator; and assigning a class to the mesh primitive on the basis of the estimated probability.
EP04709288A 2003-02-18 2004-02-09 Image segmentation by assigning classes to adaptive mesh primitives Expired - Lifetime EP1599833B1 (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
EP04709288A EP1599833B1 (en) 2003-02-18 2004-02-09 Image segmentation by assigning classes to adaptive mesh primitives

Applications Claiming Priority (4)

Application Number Priority Date Filing Date Title
EP03100371 2003-02-18
EP03100371 2003-02-18
PCT/IB2004/000394 WO2004075112A1 (en) 2003-02-18 2004-02-09 Image segmentation by assigning classes to adaptive mesh primitives
EP04709288A EP1599833B1 (en) 2003-02-18 2004-02-09 Image segmentation by assigning classes to adaptive mesh primitives

Publications (2)

Publication Number Publication Date
EP1599833A1 true EP1599833A1 (en) 2005-11-30
EP1599833B1 EP1599833B1 (en) 2007-03-21

Family

ID=32892947

Family Applications (1)

Application Number Title Priority Date Filing Date
EP04709288A Expired - Lifetime EP1599833B1 (en) 2003-02-18 2004-02-09 Image segmentation by assigning classes to adaptive mesh primitives

Country Status (6)

Country Link
US (1) US8805073B2 (en)
EP (1) EP1599833B1 (en)
JP (1) JP2006518072A (en)
AT (1) ATE357706T1 (en)
DE (1) DE602004005428T2 (en)
WO (1) WO2004075112A1 (en)

Families Citing this family (12)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
DE10357206B4 (en) * 2003-12-08 2005-11-03 Siemens Ag Method and image processing system for the segmentation of sectional image data
EP1775685B1 (en) * 2004-08-02 2019-07-17 National University Corporation Nara Institute of Science and Technology Information processing device and program
US7920730B2 (en) * 2005-10-07 2011-04-05 Siemens Medical Solutions Usa, Inc. Automatic bone detection in MRI images
US8098909B2 (en) 2007-08-31 2012-01-17 Computerized Medical Systems, Inc. Method and apparatus for efficient three-dimensional contouring of medical images
US8583209B2 (en) * 2007-10-03 2013-11-12 Siemens Aktiengesellschaft Method and system for monitoring cardiac function of a patient during a magnetic resonance imaging (MRI) procedure
US8265356B2 (en) * 2008-01-30 2012-09-11 Computerized Medical Systems, Inc. Method and apparatus for efficient automated re-contouring of four-dimensional medical imagery using surface displacement fields
US10178396B2 (en) 2009-09-04 2019-01-08 Stmicroelectronics International N.V. Object tracking
JP5542454B2 (en) * 2010-01-14 2014-07-09 株式会社東芝 Ultrasonic diagnostic apparatus, ultrasonic image processing apparatus, and ultrasonic image processing program
US8867806B2 (en) 2011-08-01 2014-10-21 Impac Medical Systems, Inc. Method and apparatus for correction of errors in surfaces
KR102070427B1 (en) * 2012-08-08 2020-01-28 삼성전자주식회사 Method and Apparatus for tracking the position of tumor
WO2017013019A1 (en) * 2015-07-17 2017-01-26 Koninklijke Philips N.V. Guidance for lung cancer radiation
CN111060520B (en) * 2019-12-30 2021-10-29 歌尔股份有限公司 Product defect detection method, device and system

Family Cites Families (23)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US4918633A (en) * 1985-11-25 1990-04-17 Eastman Kodak Company Digital image noise reduction method and transmission system
US5273040A (en) * 1991-11-14 1993-12-28 Picker International, Inc. Measurement of vetricle volumes with cardiac MRI
US5570430A (en) * 1994-05-31 1996-10-29 University Of Washington Method for determining the contour of an in vivo organ using multiple image frames of the organ
US6295464B1 (en) * 1995-06-16 2001-09-25 Dimitri Metaxas Apparatus and method for dynamic modeling of an object
JP3679512B2 (en) * 1996-07-05 2005-08-03 キヤノン株式会社 Image extraction apparatus and method
US6404920B1 (en) * 1996-09-09 2002-06-11 Hsu Shin-Yi System for generalizing objects and features in an image
US7194117B2 (en) * 1999-06-29 2007-03-20 The Research Foundation Of State University Of New York System and method for performing a three-dimensional virtual examination of objects, such as internal organs
US6466687B1 (en) * 1997-02-12 2002-10-15 The University Of Iowa Research Foundation Method and apparatus for analyzing CT images to determine the presence of pulmonary tissue pathology
US5859891A (en) * 1997-03-07 1999-01-12 Hibbard; Lyn Autosegmentation/autocontouring system and method for use with three-dimensional radiation therapy treatment planning
US6124864A (en) * 1997-04-07 2000-09-26 Synapix, Inc. Adaptive modeling and segmentation of visual image streams
US6078680A (en) * 1997-07-25 2000-06-20 Arch Development Corporation Method, apparatus, and storage medium for detection of nodules in biological tissue using wavelet snakes to characterize features in radiographic images
US6201543B1 (en) * 1997-12-17 2001-03-13 Siemens Corporate Research, Inc. Framework for segmentation of cylindrical structures using two dimensional hybrid models
DE69908526T2 (en) * 1998-04-03 2004-05-06 Koninklijke Philips Electronics N.V. Image processing method and system using contour detection steps
US6285372B1 (en) * 1998-05-08 2001-09-04 Lawrence C. Cowsar Multiresolution adaptive parameterization of surfaces
US6496184B1 (en) * 1998-11-30 2002-12-17 William T. Freeman Method for inferring scenes from test images and training data using probability propagation in a markov network
US6456287B1 (en) * 1999-02-03 2002-09-24 Isurftv Method and apparatus for 3D model creation based on 2D images
GB9906420D0 (en) * 1999-03-19 1999-05-12 Isis Innovation Method and apparatus for image processing
JP2001285894A (en) * 2000-03-31 2001-10-12 Olympus Optical Co Ltd Method for running three-dimensional image data
US6829384B2 (en) * 2001-02-28 2004-12-07 Carnegie Mellon University Object finder for photographic images
JP2004536374A (en) * 2001-03-09 2004-12-02 コーニンクレッカ　フィリップス　エレクトロニクス　エヌ　ヴィ Image segmentation
US7466848B2 (en) * 2002-12-13 2008-12-16 Rutgers, The State University Of New Jersey Method and apparatus for automatically detecting breast lesions and tumors in images
US7499572B2 (en) * 2003-06-30 2009-03-03 The Salk Institute For Biological Studies Surface segmentation from luminance and color differences
GB2409028A (en) * 2003-12-11 2005-06-15 Sony Uk Ltd Face detection

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
See references of WO2004075112A1 *

Also Published As

Publication number Publication date
EP1599833B1 (en) 2007-03-21
ATE357706T1 (en) 2007-04-15
DE602004005428D1 (en) 2007-05-03
JP2006518072A (en) 2006-08-03
US8805073B2 (en) 2014-08-12
WO2004075112A1 (en) 2004-09-02
DE602004005428T2 (en) 2007-11-29
US20070041639A1 (en) 2007-02-22

Similar Documents

Publication Publication Date Title
US7680312B2 (en) Method for knowledge based image segmentation using shape models
JP6626344B2 (en) Image processing apparatus, control method for image processing apparatus, and program
AU768446B2 (en) System and method for 4D reconstruction and visualization
US8131038B2 (en) System and method for global-to-local shape matching for automatic liver segmentation in medical imaging
JP6505124B2 (en) Automatic contour extraction system and method in adaptive radiation therapy
JP2020503603A (en) System and method for stochastic segmentation in anatomical image processing
US20050238233A1 (en) Computation of contour
EP1599833B1 (en) Image segmentation by assigning classes to adaptive mesh primitives
WO2006024974A1 (en) Feature weighted medical object contouring using distance coordinates
US20060210158A1 (en) Object-specific segmentation
US20060093197A1 (en) Semi-local active contour for a faint region detection in knee CT images
EP3373247A1 (en) Image segmentation and prediction of segmentation
Wang et al. Magnetic resonance image analysis by information theoretic criteria and stochastic site models
US20090161926A1 (en) Semi-automatic Segmentation of Cardiac Ultrasound Images using a Dynamic Model of the Left Ventricle
Fritz et al. Segmentation of the left and right cardiac ventricle using a combined bi-temporal statistical model
US8144987B2 (en) Method, a system and a computer program for segmenting a surface in a multi-dimensional dataset
JP6827707B2 (en) Information processing equipment and information processing system
CN115439423B (en) CT image-based identification method, device, equipment and storage medium
van Ginneken et al. Interactive shape models
WO2006110398A2 (en) Estimation of solitary pulmonary nodule diameters with a hybrid segmentation approach
Klemencic et al. Non-rigid registration based active appearance models for 3D medical image segmentation
JP5955782B2 (en) Image analysis
Hoogendoorn et al. A groupwise mutual information metric for cost efficient selection of a suitable reference in cardiac computational atlas construction
Desrosiers An unsupervised random walk approach for the segmentation of brain MRI
JP7356330B2 (en) Image processing device, image processing method, and program

Legal Events

Date Code Title Description
PUAI Public reference made under article 153(3) epc to a published international application that has entered the european phase

Free format text: ORIGINAL CODE: 0009012

17P Request for examination filed

Effective date: 20050919

AK Designated contracting states

Kind code of ref document: A1

Designated state(s): AT BE BG CH CY CZ DE DK EE ES FI FR GB GR HU IE IT LI LU MC NL PT RO SE SI SK TR

AX Request for extension of the european patent

Extension state: AL LT LV MK

DAX Request for extension of the european patent (deleted)
GRAP Despatch of communication of intention to grant a patent

Free format text: ORIGINAL CODE: EPIDOSNIGR1

RAP1 Party data changed (applicant data changed or rights of an application transferred)

Owner name: KONINKLIJKE PHILIPS ELECTRONICS N.V.

Owner name: PHILIPS INTELLECTUAL PROPERTY & STANDARDS GMBH

GRAS Grant fee paid

Free format text: ORIGINAL CODE: EPIDOSNIGR3

GRAA (expected) grant

Free format text: ORIGINAL CODE: 0009210

AK Designated contracting states

Kind code of ref document: B1

Designated state(s): AT BE BG CH CY CZ DE DK EE ES FI FR GB GR HU IE IT LI LU MC NL PT RO SE SI SK TR

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: FI

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20070321

Ref country code: AT

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20070321

Ref country code: NL

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20070321

Ref country code: BE

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20070321

Ref country code: LI

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20070321

Ref country code: CH

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20070321

Ref country code: SI

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20070321

REG Reference to a national code

Ref country code: GB

Ref legal event code: FG4D

RIN1 Information on inventor provided before grant (corrected)

Inventor name: KAUS, MICHAELPHILIPS I. P. & STANDARDS GMBH

Inventor name: VON BERG, JENSPHILIPS I.. P. & STANDARDS GMBH

REG Reference to a national code

Ref country code: CH

Ref legal event code: EP

REF Corresponds to:

Ref document number: 602004005428

Country of ref document: DE

Date of ref document: 20070503

Kind code of ref document: P

REG Reference to a national code

Ref country code: IE

Ref legal event code: FG4D

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: SE

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20070621

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: ES

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20070702

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: PT

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20070821

REG Reference to a national code

Ref country code: CH

Ref legal event code: PL

NLV1 Nl: lapsed or annulled due to failure to fulfill the requirements of art. 29p and 29m of the patents act
EN Fr: translation not filed
PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: SK

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20070321

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: RO

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20070321

Ref country code: CZ

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20070321

PLBE No opposition filed within time limit

Free format text: ORIGINAL CODE: 0009261

STAA Information on the status of an ep patent application or granted ep patent

Free format text: STATUS: NO OPPOSITION FILED WITHIN TIME LIMIT

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: DK

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20070321

26N No opposition filed

Effective date: 20071227

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: GR

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20070622

Ref country code: FR

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20071123

Ref country code: IT

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20070321

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: MC

Free format text: LAPSE BECAUSE OF NON-PAYMENT OF DUE FEES

Effective date: 20080228

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: FR

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20070321

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: IE

Free format text: LAPSE BECAUSE OF NON-PAYMENT OF DUE FEES

Effective date: 20080211

Ref country code: EE

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20070321

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: CY

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20070321

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: BG

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20070621

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: HU

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20070922

Ref country code: LU

Free format text: LAPSE BECAUSE OF NON-PAYMENT OF DUE FEES

Effective date: 20080209

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: TR

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20070321

REG Reference to a national code

Ref country code: DE

Ref legal event code: R081

Ref document number: 602004005428

Country of ref document: DE

Owner name: PHILIPS GMBH, DE

Free format text: FORMER OWNER: PHILIPS INTELLECTUAL PROPERTY & STANDARDS GMBH, 20099 HAMBURG, DE

Effective date: 20140327

Ref country code: DE

Ref legal event code: R081

Ref document number: 602004005428

Country of ref document: DE

Owner name: PHILIPS DEUTSCHLAND GMBH, DE

Free format text: FORMER OWNER: PHILIPS INTELLECTUAL PROPERTY & STANDARDS GMBH, 20099 HAMBURG, DE

Effective date: 20140327

Ref country code: DE

Ref legal event code: R082

Ref document number: 602004005428

Country of ref document: DE

Representative=s name: MEISSNER, BOLTE & PARTNER GBR, DE

Effective date: 20140327

Ref country code: DE

Ref legal event code: R082

Ref document number: 602004005428

Country of ref document: DE

Representative=s name: MEISSNER BOLTE PATENTANWAELTE RECHTSANWAELTE P, DE

Effective date: 20140327

REG Reference to a national code

Ref country code: DE

Ref legal event code: R082

Ref document number: 602004005428

Country of ref document: DE

Representative=s name: MEISSNER, BOLTE & PARTNER GBR, DE

Ref country code: DE

Ref legal event code: R081

Ref document number: 602004005428

Country of ref document: DE

Owner name: PHILIPS GMBH, DE

Free format text: FORMER OWNER: PHILIPS DEUTSCHLAND GMBH, 20099 HAMBURG, DE

Ref country code: DE

Ref legal event code: R082

Ref document number: 602004005428

Country of ref document: DE

Representative=s name: MEISSNER BOLTE PATENTANWAELTE RECHTSANWAELTE P, DE

PGFP Annual fee paid to national office [announced via postgrant information from national office to epo]

Ref country code: GB

Payment date: 20170228

Year of fee payment: 14

PGFP Annual fee paid to national office [announced via postgrant information from national office to epo]

Ref country code: DE

Payment date: 20170428

Year of fee payment: 14

REG Reference to a national code

Ref country code: DE

Ref legal event code: R119

Ref document number: 602004005428

Country of ref document: DE

GBPC Gb: european patent ceased through non-payment of renewal fee

Effective date: 20180209

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: DE

Free format text: LAPSE BECAUSE OF NON-PAYMENT OF DUE FEES

Effective date: 20180901

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: GB

Free format text: LAPSE BECAUSE OF NON-PAYMENT OF DUE FEES

Effective date: 20180209